Given a demo relation (a pair of word forms) and a query word, we devise a character-based recurrent neural network architecture using three separate encoders and a decoder, trained to predict the missing second form of the query word. Our results show that the exact form can be predicted for English with an accuracy of 94.7%. For Swedish, which has a more complex morphology with more inflectional patterns for nouns and verbs, the accuracy is 89.3%.
To appear at Subword & Character Level Models in NLP (SCLeM) workshop at EMNLP 2017 in Copenhagen, Denmark, September 7.
Olof Mogren, Richard Johansson
Generative adversarial networks have been proposed as a way of efficiently training deep generative neural networks. We propose a generative adversarial model that works on continuous sequential data, and apply it by training it on a collection of classical music. We conclude that it generates music that sounds better and better as the model is trained, report statistics on generated music, and let the reader judge the quality by downloading the generated songs.
Constructive Machine Learning Workshop (CML) at NIPS 2016 in Barcelona, Spain, December 10.
The EMNLP conference took place in Copenhagen in September 2017. In this blog post I share some observations that I made during the conference. These included subword-level models, multilingual NLP, language grounding, and inspiration from children.
August 7-12, the 54th conference of the Association of Computational Linguistics (ACL) took place at the Humboldt University in Berlin. This blog post contains a write-up of some of my favourite presentations during the conference.
On March 23rd, 2018 at 10:00, I successfully defended my doctorate thesis titled
“Representation learning for natural language”.
On November 20th, 2015 at 10:00, I successfully defended my licentiate thesis titled
“Multi-document summarization and semantic relatedness”.
Discussion leader was
Tapani Raiko from Aalto University.
The following students recently wrote their master's theses under my supervision.
Christian Meijner, Simon Persson:
Blood Glucose Prediction for Type 1 Diabetes using Machine Learning
Long Short-term Memory based models for blood glucose prediction with confidence level outputs.
2017-05-14: Can we trust AI: A talk at the science festival
During the science festival in Gothenburg, we had a session discussing artificial intelligence. The theme for the whole festival was “trust”, so we naturally named our session “Can we trust AI”. I gave an introduction, and shared my view of some of the recent progress that has been made in AI and machine learning, and then we had four other speakers giving their views of current state of the art. Finally, I chaired a discussion session that was much appreciated with the audience. The room was filled, and many people came up to us afterwards and kept the discussion going. The other speakers were Annika Larsson from Autoliv, Ola Gustavsson from Dagens Nyheter, and Hans Salomonsson from Data Intelligence Sweden AB.
2017-02-02: Takeaways from NIPS: meta-learning and one-shot learning
(Chalmers Machine Learning Seminars)
Before the representation learning revolution, hand-crafted features were a prerequisite for a successful application of most machine learning algorithms. Just like learned features have been massively successful in many applications, some recent work has shown that you can also automate the learning algorithms themselves. In this talk, I'll cover some of the related ideas presented at this year's NIPS conference.
2016-10-06: Deep Learning Guest Lecture
(FFR135, Artificial Neural Networks)
A motivational talk about deep artificial neural networks, given to the students in FFR135 (Artificial neural networks). I gave motivations for using deep architechtures, and to learn hierarchical representations for data.
2016-09-29: Recent Advances in Neural Machine Translation
(Chalmers Machine Learning Seminars)
Neural models for machine translation was introduced seriously in 2014. With the introduction of attention models their performance improved to levels comparable to those of statistical phrase-based machine translation, the type of translation we are all familiar with through servies like Google Translate.
However, the models have struggled with problems like limited vocabularies, the need of large amounts of data for training, and that they are expensive to train and use.
In the recent months, a number of papers have been published to remedy some of these issues. This includes techniques to battle the limited vocabulary problem, and of using monolingual data to improve the performance. As recently as Monday evening (Sept 26), Google uploaded a paper on their implementation of these ideas, where they claim performance on par with human translators, both counted in BLEU scores, and in human evaluations.
During this talk, I'll go through the ideas behind these recent papers.
March 23, 2018, I defended my PhD thesis, Representation learning for natural language (click for more info).
During 2016-2017, I was the organizer of Chalmers machine learning seminars.
In 2016, I taught a PhD course in Deep Learning, together with Mikael Kågebäck and Fredrik Johansson. I have also taught Algorithms for Machine Learning and Inference, AI (specifically the parts about probabilistic methods, including probabilistic graphical models), Object Oriented Programming, Data Structures, and Algorithms (basic course, and advanced course).